Confabulation what is associated with its rise and fall a study in brain injury

onfabulation what is associated with its rise and fall a study in brain injury

1. Special issue: Research report
Confabulation: What is associated with its rise and
fall? A study in brain injury
Ana Bajo
a,c,*
, Simon Fleminger
a,c
, Chris Metcalfe
b
and
Michael D. Kopelman
a
a
Department of Psychological Medicine, King's College London, Institute of Psychiatry, Psychology and Neuroscience
(IoPPN), Academic Unit of Psychiatry, London, UK
b
School of Social and Community Medicine, University of Bristol, Bristol, UK
c
Brain Injury Rehabilitation Unit, Edgware Hospital, Middlesex, UK
article info
Article history:
Received 17 March 2016
Reviewed 1 May 2016
Revised 3 June 2016
Accepted 10 June 2016
Published online 27 June 2016
Keywords:
Confabulation
Brain injury
Memory
Mood
Cognition
abstract
The aim of this study was to investigate cognitive and emotional factors associated with
the presence and clinical course of confabulation.
24 confabulating participants were compared with 11 brain injured and 6 healthy
controls on measures of temporal context confusions (TCC), mood state (elation, depres-
sion) and lack of insight. Measures of autobiographical memory and executive function
were also available. Changes in confabulation and these other measures were monitored
over 9 months in the confabulating group.
We found that TCC were more common in confabulating patients than in healthy
controls, and that the decline in these errors paralleled the recovery from confabulation.
However, TCC were not specific to the presence of confabulation in brain injury; and their
decline was not correlated with change in confabulation scores over 9 months. We found
that elated mood and lack of insight discriminated between confabulating and non-
confabulating patients, but these measures did not correlate with either the severity of
confabulation or change in confabulation scores through time. What seems to have been
most strongly associated with the severity of confabulation scores at ‘baseline ’and
changes through time (over 9 months) were the severity of memory impairment (especially
on autobiographical memory) and errors on executive tests (particularly in making
cognitive estimates). Greater autobiographical memory and executive impairment were
associated with more severe confabulation.
Abbreviations:AMI, autobiographical memory interview; ACoA, anterior communicating artery aneurysm; CET, cognitive estimation
test; DK, don't know; IQ, intelligence quotient; MREC, Multicentre Research Ethics Committee (UK); MTL, medial temporal lobe; NART-R,
National Adult Reading TesteRevised; PTA, post traumatic amnesia; ROC, receiver operating characteristics; SCOLP, speed and capacity
of language processing; T1, baseline assessment (Time 1); T3, 9-month follow up (Time 3); TMT, trail making test; WAIS-III, Wechsler
Intelligence ScaleeIII; WMS-III, Wechsler Memory ScaleeIII; TCC, temporal context confusion; WASI, Wechsler Abbreviated Scale of
Intelligence.
*Corresponding author. Present address: Department of Old Age Psychiatry, King's College London (IoPPN), De Crespigny Park, London
SE5 8AF, UK.
E-mail addresses:[email protected](A. Bajo),s.fl[email protected](S. Fleminger),[email protected](C. Metcalfe),
[email protected](M.D. Kopelman).
Available online atwww.sciencedirect.com
ScienceDirect
Journal homepage:www.elsevier.com/locate/cortex
cortex 87 (2017) 31 e43
http://dx.doi.org/10.1016/j.cortex.2016.06.016
0010-9452/©2016 Elsevier Ltd. All rights reserved.

2. The findings were consistent with the view that confabulation results from executive
dysfunction where autobiographical memory is also impaired; and that it resolves as these
impairments subside.
©2016 Elsevier Ltd. All rights reserved.
1. Introduction
Confabulations are false or erroneous memories arising
involuntarily in the context of a neurological amnesia. Their
content may be entirely or substantially erroneous, or they
may consist of real memories jumbled up and retrieved out of
context (Kopelman, 2010). Confabulated memories are usually
autobiographical; they arise unintentionally; and the patient
is usually unaware of his/her condition (Burgess&Shallice,
1996). Confabulations often compel the person to act upon
them (Schnider, 2001). Confabulation is a very debilitating
condition: the person may self-discharge from hospital
against medical advice, and social support networks are often
strained. Relatives may suffer unjustified paranoid accusa-
tions, and/or struggle to appreciate the unintentional nature
of the confabulations (Baddeley&Wilson, 1986). Several
theoretical models have attempted to explain the underlying
cognitive deficits which give rise to confabulation. However,
to date there is no consensus regarding the definition, classi-
fication, or causes of this clinical condition, how it arises, and
its clinical course.
Spontaneous confabulation has been defined as“a persis-
tent, unprovoked outpouring of erroneous memories ”
(Kopelman, 1987, 1999, 2010), whereas momentary or pro-
voked confabulations consist of fleeting intrusion errors or
distortions when memory is challenged or tested. However,
this distinction has been challenged. On the one hand,
Schnider and colleagues (Schnider, 2003, 2008; Schnider, von
Daniken,&Gutbrod, 1996) have argued that ‘spontaneous
confabulation’should be reserved for patients who act upon
their false memories; and they have argued for other varieties
of confabulation. On the other hand,Gilboa and Moscovitch
(2002), Gilboa et al. (2006)andMcVittie, McKinlay, Della Sala,
and MacPherson (2014)have argued that there is overlap be-
tween these categories of confabulation or, at least, not a clear
boundary between them (very florid confabulations can be, at
least partially, ‘provoked’). Nevertheless, there appears to be
considerable agreement that spontaneous confabulation ari-
ses when executive dysfunction and frontal lobe damage are
superimposed upon an amnesic deficit (Baddeley&Wilson,
1986; Kapur&Coughlan, 1980; Luria, 1976; Stuss, Alexander,
Lieberman,&Levine, 1978). Moreover, more recent neuro-
imaging and evoked potential investigations have particularly
pinpointed damage within the ventro-medial and/or orbito-
frontal brain regions (Gilboa&Verfaellie, 2010; Gilboa et al.,
2006; Hebscher, Barkan-Abramski, Goldsmith, Aharon-
Peretz,&Gilboa, 2015; Manuel, David, Bikson,&Schnider,
2014; Nahum, Bouzerda-Wahlen, Guggisberg, Ptak, &
Schnider, 2012; Schnider, 2003; Schnider et al., 2013; Toosy
et al., 2008; Turner, Cipolotti, Yousry,&Shallice, 2008).
There are various theories of confabulation, which have
been reviewed elsewhere (DeLuca, 2009; Fotopoulou, 2010;
Kopelman, 1999, 2010; Schnider, 2008). These include: (i) the-
ories of context memory confusion or a deficit in ‘reality
monitoring’(Schnider, 2003, 2008); (ii) the theory that tem-
poral consciousness is intact, but malfunctioning (Dalla Barba,
1993; Dalla Barba, Cappelletti, Signorini,&Denes, 1997; La
Corte, Serra, Attali, Boisse,&Dalla Barba, 2010); (iii) theories
which emphasise deficits in strategic retrieval, trace specifi-
cation, and verification/editing processes (Burgess&
Shallice,
1996; Gilboa et al., 2006; Schacter, Norman,&Koutstaal, 1998);
and (iv) motivational accounts of confabulation (Conway&
Tacchi, 1996; Fotopoulou, Conway, Griffiths, Birchall, &
Tyrer, 2007; Fotopoulou, Conway, Solms, Tyrer,&Kopelman,
2008). Finally, (v) interactionist/multifactorial accounts
combine elements of several of these theories ( Johnson,
O'Connor,&Cantor, 1997; Kopelman, Ng,&Van Den Brouke,
1997).
Of particular relevance to this study is Schnider's account
because it has attempted to explain not only the cause of
behaviourally spontaneous confabulations, but also their
clinical course (Nahum, Ptak, Leemann, Lalive,&Schnider,
2010; Schnider, 2008; Schnider, Ptak, von Daniken, &
Remonda, 2000). In a series of elegant experiments, Schnider
and colleagues compared spontaneous confabulators to non-
confabulating amnesic patients (Schnider&Ptak, 1999;
Schnider et al., 1996, 2000). They found only one feature
distinguished between the two groups: performance on a
continuous recognition task at a second ‘run’, where 8 previ-
ous targets had become distractors and 8 distractors now
became targets presented 6 times, a potentially very useful
diagnostic task. The authors inferred that the key cognitive
deficit giving rise to confabulation was temporal context
confusion (TCC). In an important study,Schnider et al. (2000)
found that performance on the TCC task improved as pa-
tients recovered from confabulation, whereas other cognitive
measures did not change in parallel with recovery from
confabulation. Subsequently,Schnider (2003, 2008)reinter-
preted TCC as reflecting a defective ‘filtering’mechanism with
respect to incoming information or ‘ongoing reality’. More
specifically, they postulated a deficit in suppressing memories
that no longer corresponded to the present.Schnider (2008)
explained that this was why confabulators appear to ‘live in
the past’, and are so convinced of their false memories that
they act upon them.Schnider (2008)distinguished between
four types of confabulation: intrusions (provoked confabula-
tions), momentary, fantastic, and behaviourally spontaneous
confabulations.Nahum et al. (2012)andBouzerda-Wahlen,
Nahum, Ptak, and Schnider (2013)have argued that, while
momentary confabulations are associated with a number of
cortex 87 (2017) 31 e4332

3. cognitive impairments, including executive deficits, behav-
iourally spontaneous confabulation and disorientation are
associated with deficient reality monitoring, an inability to
learn that one stimulus no longer predicts the occurrence of
another. This was attributed to a defective orbito-frontal
filtering mechanism.
Gilboa et al. (2006)used the TCC task to explore these
findings. They compared a group of confabulating patients
who had suffered anterior communicating artery aneurysm
(ACoA), non-confabulating patients with similar underlying
pathology, amnesic patients with medial temporal lobe (MTL)
lesions, and healthy controls. Consistent withSchnider et al.'s
(2000, 1996)results, confabulating patients'performance on
the TCC task was worse than that of healthy controls and
amnesic (MTL) patients. However, the task did not discrimi-
nate between confabulating and non-confabulating patients
with ACoA and those without. The non-confabulating ACoA
group overlapped with both the confabulating patients and
the healthy controls. Moreover,Gilboa et al. (2006)obtained a
similar pattern of results on a task in which participants were
asked to make difficult, visuo-perceptual discriminations,
instead of temporal context judgements.Gilboa et al. (2006)
concluded that, whilst TCC errors might be sensitive to
confabulation, they were not specific to this condition. How-
ever, it should be noted that there were only five confabulators
in the originalSchnider et al. (1996)investigation, and only
four confabulators inGilboa et al.'s (2006)study.
Other theorists have attributed spontaneous confabulation
to problems in trace specification or verification at retrieval
(Kopelman, 1999, 2010). For example,Burgess and Shallice
(1996)stated that there were deficits either in ‘strategic con-
trol’, or trace specification, or the editing out of errors, and
Moscovitch and Melo (1997) andSchacter et al. (1998)
employed different terminology to argue very similarly.
Gilboa et al. (2006)placed emphasis upon impaired strategic
retrieval and, in particular, defective pre-conscious, post-
retrieval monitoring (Gilboa&Verfaellie, 2010; Hebscher et al.,
2015). More recently,Ghosh, Moscovitch, Colella, and Gilboa
(2014)have argued that patients with ventro-medial prefron-
tal cortical lesions have difficulty instantiating task-relevant
schemas, and, instead, rely on inappropriately activated
memory schemas (Ghosh&Gilboa, 2014) even in non-memory
tasks. These authors demonstrated impaired reaction times
by confabulating patients in processing schema and in
rejecting inappropriate schema, even though their reaction
time on other tasks was not impaired.
By contrast, the motivational account of confabulation has
focussed upon cognitive factors and the influence of mood
and emotion (Fotopoulou, 2008; Fotopoulou, Solms, &
Turnbull, 2004; Metcalf, Langdon, &Coltheart, 2010;
Turnbull, Jenkins,&Rowley, 2004). Several authors have
found that confabulations may promote the individual's self-
perception in a favourable or idealised light (Fotopoulou,
2010; Turnbull, Berry,&Evans, 2004). In this way, confabula-
tions might help the individual cope with a rather depressing
reality (e.g. being in hospital, loss of job/status, etc.).
Furthermore, there is some indication that confabulations
may be related to depressed mood, although more research is
required (Fotopoulou, Conway, Solms et al., 2008; Fotopoulou,
Conway, Tyrer, et al., 2008; Metcalf et al., 2010). In a previous
study (Bajo, Fleminger,&Kopelman, 2010), we found that the
content of confabulations can be, but is not always, associated
with strong emotion; and that this emotions may be either
pleasant or unpleasant. One of the aims of the present study
was to explore whether mood state (depression; elation) is
associated with confabulation.
There is evidence from several studies of psychiatric pa-
tients that elated mood affects memory (Corwin, Peselow,
Feenan, Rotrosen,&Fieve, 1990; Gainotti&Marra, 1994).
Corwin et al. (1990)found that, during a recognition memory
task, patients with elated mood (or ‘hypomania’) gave many
more false positives than controls, demonstrating a liberal
response bias. This subsided when their mood disorder was
treated with beta-blockers, the patients then adopting a more
conservative decision-making bias. There is also clinical evi-
dence suggesting that confabulators may show mild elation
and/or poor insight (Weinstein&Kahn, 1950). Lack of insight
is also considered an integral characteristic of the syndrome
of confabulation in some theoretical models of confabulation
(e.g.Burgess&Shallice, 1996). Furthermore evidence suggests
that recovery from confabulation may be accompanied by
increased awareness of amnesia and confabulation (DeLuca,
2000; Feinberg&Roane, 1997; Mercer, Wapner, Gardner,&
Benson, 1977; Nahum et al., 2010). However, elated mood
and insight have not been formally investigated in previous
studies of confabulation. Consequently, in this study, we
explored whether elated mood and/or limited insight were
related to confabulation.
In summary, many authors have viewed confabulation as
the consequence of cognitive deficits, such as in temporal
context memory or reality monitoring, or alternatively stra-
tegic retrieval, trace specification, and/or editing/verification
processes. In addition, there is evidence indicating that con-
fabulations may also be modulated by motivational factors,
including mood state and lack of insight. Elated mood in
psychiatric patients leads to significant false recognition on
memory tasks, but this improves when the underlying mood
disorder is successfully treated.
In the present investigation, we studied a large group of
confabulating patients, non-confabulating brain-injured pa-
tients, and healthy controls to attempt to resolve the following
issues:
1) Are TCC predictive of the presence of confabulation and
its time course (the Schnider hypothesis?)
2) Are affective factors (mood state), or a lack of ‘insight’,
predictive of confabulation and its time-course (as
Corwin et al.'s 1991findings suggested)?
3) Are memory and executive performance predictive of
confabulation and its time-course (the ‘traditional’
notion)?
2. Materials and method
2.1. Participants
Written consent was obtained for all participants according to
procedures approved by a Multicentre Research Ethics
cortex 87 (2017) 31 e43 33

4. Committee (MREC). Three groups of participants were
recruited: (i) 24 confabulating brain injured patients; (ii) 11
non-confabulating brain injured patients, and (iii) six healthy
participants.
In order to be included in the study patients were required
to have a primary diagnosis of acquired brain injury resulting
from pathologies such as traumatic brain injury, hypoxia,
subarachnoid haemorrhage, Wernicke-Korsakoff syndrome,
cerebral infection and brain tumour, each diagnosed by the
responsible clinician accordingly. The confabulating patients
were selected on the basis of clinical evidence of spontaneous
confabulationanda score of 8 or above on the ‘episodic’
component ofDalla Barba's (1993)Confabulation Battery (see
below) (Kopelman et al., 1997). 20 out of 24 of the confabula-
tors had acted upon their confabulations (e.g. a school teacher
who patrolled the hospital corridors, believing he was
inspecting the dormitories at a boarding school; another pa-
tient who travelled 30 miles to his home, believing he was still
living there). Clinical neuroimaging (MRI, CT) was available for
21 of the patients: 10 showed ventro-medial prefrontal or
orbito-frontal pathology, and eight showed atrophy encom-
passing these brain regions, whereas three showed no visible
changes at these sites. Patients withnoclinical diagnosis of
confabulation and a score of 4 or less on the episodic section of
the Confabulation Battery were included in the brain injured
non-confabulating group.
2.1.1. Recruitment sites
Patients were recruited via consecutive referrals mainly from
in-patient, but also from out-patient clinics of the following
hospitals: Edgware Hospital Brain Injury Rehabilitation Unit
(BIRU), South London and the Maudsley Trust (SLAM), Guy's
and St Thomas'Trust hospitals, and Blackheath Brain Injury
Rehabilitation Unit. Staff at these sites were asked to refer to
us any patient they thought was confabulating. In order to
recruit healthy controls, relatives or carers of brain injured
individuals were approached at BIRU's outpatient clinic.
2.1.2. Ascertainment
2.1.2.1. C
ONFABULATING PATIENTS .A total of 46 confabulating pa-
tients were referred to us and assessed. Three of these refused
to participate in the study. Two patients started the study but
did not complete the follow-ups. One patient was excluded
because severe confusion prevented meaningful conversation.
Another patient was suffering a major mental health illness at
the time of recruitment and was also excluded. Five patients
were excluded because no informant was available to verify
their confabulations. Another 10 patients were interviewed and
excluded because they did not meet the criteria for evidence of
confabulatory behaviour set above. This left 24 confabulating
brain injured participants in the experimental group.
2.1.2.2. NON-CONFABULATING PATIENTS .Twenty three non-
confabulating brain injured controls were approached. Two
of these refused to participate in this study. Another 2 did not
complete the follow-ups. One patient was excluded because
no informer was available to verify their memories. A further 7
patients were excluded because they did not meet the criteria
for matching them to the experimental group. This left 11
non-confabulating brain injured patients in the control group;
they were matched to confabulators for severity of memory
disorder (using the WMS delayed memory subtest) and level of
education (Schnider et al., 1996), (seeTable 1for details); they
were also matched for severity of cognitive deficits (according
to MMSE scores) and age.
2.1.2.3. HEALTHY PARTICIPANTS.Six healthy participants were
approached and recruited. They were matched to the experi-
mental group on age and level of education.
2.2. Background assessments
The following background neuropsychological tests were
carried out.
2.2.1. General reasoning tests
- the revised National Adult Reading Test (NART-R) as a
measure of premorbid IQ (Nelson&Willison, 1991); the
Wechsler Abbreviated Scale of Intelligence (WASI) or the
Wechsler Adult Scale of Intelligence-III (WAIS-III) as mea-
sures of current IQ (Wechsler, 1999; Wechsler, Wycherley,
Benjamin, Crawford,&Mockler, 1997a);
2.2.2. Speed of processing
- the speed and capacity of language processing (SCOLP) (A.
D.Baddeley, Emslie,&Nimmo-Smith, 1992) and the WAIS-
III ‘digit symbol’subtest (Wechsler, et al., 1997a);
2.2.3. Executive function
- the trail making test (TMT) (Reitan, 1958), the cognitive
estimates test (CET) (Shallice&Evans, 1978), and the Hay-
ling&Brixton Tests (Burgess&Shallice, 1997).
2.2.4. Memory
- the Wechsler Memory Scale-III as a measure of antero-
grade memory (Wechsler, Wycherley, Benjamin, Crawford,
&Mockler, 1997b). The childhood, adulthood and recent
episodic events sections of the autobiographical memory
interview (AMI) (Kopelman, Wilson,&Baddeley, 1990) was
employed as a measure of retrograde personal memory.
2.3. Procedures
The experimental tasks were carried out on inclusion to the
study. In the case of confabulating patients, this was at a
mean of 6 months following their brain injury (Table 1). These
tests were repeated 3 months later, and again 9 months after
the initial assessment. The tasks were:
2.3.1. TheDalla Barba (1993)Confabulation Battery adapted
for the UK (Kopelman, et al., 1997)
This interview asks questions about orientation in time and
place, episodic memory, personal semantic memory, and
cortex 87 (2017) 31 e4334

5. items designed to provide a ‘Don't Know’response. Responses
were recorded verbatim, and these were verified against in-
formation provided by relatives or held within medical re-
cords. Healthy participants were given only the general
semantic and orientation sections of the Confabulation Bat-
tery, as it was not possible to verify their answers to the
remaining questions.
2.3.2.Schnider et al.'s (1996)TCC task
In the first ‘run’of this task, participants were shown 120
(Snodgrass&Vanderwart, 1980) line drawings of objects and
animals on a computer screen, eight drawings were repeated
six times (targets), and 112 drawings were shown only once
(distractors). Participants were asked to say whether they had
seen a given slide before or not. After a 20-min time-delay, a
second ‘run’was administered, in which previous targets
became distractors, and eight distractors were now targets,
shown six times. Participants were asked to say whether they
had seen a given item in this ‘run’or not. In order to calculate
TCC, the following formula was used:TCC¼(FP2/H2)(FP1/
H1)(FP¼false positives in run 1 or 2; H¼hits in run 1 or 2)
(Schnider et al., 1996, 2000).
2.3.3. Current mood state
- The Altman self-rating mania scale (ASRM)emania sub-
scale for elated mood (Altman, Hedeker, Peterson,&Davis,
1997) and the hospital anxiety and depression scale for
depressed mood (HADS) (Zigmond&Snaith, 1983) were
used.
2.3.4. Insight
- A semi-structured interview was constructed specifically
for this study. It tapped awareness of reasons for hospi-
talisation, of the presence of brain damage, and of deficits
in memory and thinking, and in general self-care. There
were seven items, each score 0e2 with high scores indi-
cating a lack of insight (seeSupplementary material
Appendix 1).
2.4. Statistical analyses and experimental design
Differences between the three groups (confabulators, non-
confabulating brain-injured, healthy controls) on back-
ground neuropsychological tests, mood, and cognitive mea-
sures were investigated using one way or two way ANOVAs or
t-tests. KruskaleWallis, and ManneWhitneyUtests were
used where parametric assumptions were not met. Chi-
square was used to examine frequency distributions, with a
Friedman correction for multiple comparisons as appropriate.
With parametric tests, the Holm's Bonferroni correction for
multiple comparisons was used (Walsh, 2004).
Correlational analyses (Pearson'sr) were used to examine
the associations between variables, and, in particular, the
correlates between change scores over 9 months. A backward
stepwise binary logistic regression was used as a first step for
determining the receiver operating characteristics (ROC) curve
in differentiating confabulating from non-confabulating brain
injured patients. A forward stepwise multiple regression was
used to determine the degree of variance attributable to
particular variables.
3. Results
Table 1shows age, and level of education for the three groups.
There were no significant differences between the groups in
terms of gender, age, or education.Table 1also shows the
duration of post traumatic amnesia (PTA) and time since
injury. There was a trend for these to be longer in the non-
confabulating group, but the two brain-injured groups did
not differ significantly in this regard. It should be noted that
the confabulation group was first assessed at a mean of 6
months (±3.7) following their onset, meaning that their 9-
month follow-up occurred at a mean of 15 months following
the onset of their disorder.
Table 2shows summary statistics for confabulating and
non-confabulating patients on background cognitive
reasoning and processing measures. NART-R and WASI mean
scores indicated that Full Scale I.Q. had declined by approxi-
mately 13e14 points, relative to predicted premorbid (NART-
R) scores, in both brain-injured groups at inclusion in the
study. SCOLP and Digit Symbol scores indicated reduced
speed of processing (below the 5th percentile). There were no
significant differences between the two brain-injured groups
on any of these measures.Table 2also indicates that both
brain-injured groups were severely and comparably impaired
across a range of executive tests (Trail Making B, cognitive
estimates, Hayling, Brixton). Likewise, both groups were
severely but comparably impaired at standard verbal and vi-
sual, recall and recognition anterograde memory tasks. There
were no significant differences between the two patient
groups at any of these tasks.
Table 2also shows mean accuracy scores of two brain-
injured groups on the episodic schedule of the AMI. Confab-
ulating patients scored ‘borderline abnormal ’level for
Table 1eDemographic and injury characteristics. Mean (SD) of age, years of formal education, post traumatic amnesia
(PTA), and post-injury interval of 3 groups: brain injured confabulating patients, brain injured amnesic non-confabulating
patients and healthy controls.
N Mean age Education (yrs) Post-injury (months) PTA (days)
Confabulating brain-injured 24
a
51.5 (±10.2) 12.1 ( ±2.8) 6.1 ( ±3.7) 30.4 ( ±28.4)
Non-confabulating brain-injured 11 41.9 ( ±13.3) 12.1 ( ±2.8) 11.4 ( ±11.6) 48.3 ( ±41.4)
Healthy controls 6 51.2 ( ±8.5) 13.5 ( ±2.2) ee
a
20 had acted on their confabulations. 21 had imaging available: 10¼VMF/O-F pathology; 8 widespread atrophy; 3 no visible VMF/O-F change.
cortex 87 (2017) 31 e43 35

6. childhood and young adulthood memories, and ‘definitely
abnormal’for recent personal events. Non-confabulating
brain-injured patients performed at the ‘acceptable’level for
childhood, but ‘borderline abnormal’for young adult mem-
ories, and ‘definitely abnormal’for recent autobiographical
memories. The pattern of performance in both groups was
consistent with a temporal (or ‘Ribot’) gradient, i.e. relative
sparing of early memories. A repeated measures two-way
ANOVA gave a significant main effect of group (F
2,32¼5.85,
P¼.02), indicating that the non-confabulating patients scored
more highly than the confabulators. There was also a signifi-
cant effect of time period: (F
1.70,54.44
1 ¼7.00,P¼.04). But there
was no significant group by time eperiod interaction
(F
1.70,54.44
1 ¼1.30,P¼.28), indicating the slope of the ‘temporal
gradient’did not differ between the two patient groups.
Planned pairwise comparisons showed that in both groups
recall of recent events was significantly less accurate than
recall of childhood memories (mean difference¼1.32, .05) and
of adulthood events (mean difference¼1.5,P¼.01), after
Bonferroni correction for multiple comparisons. This indi-
cated that in both groups events that happened during par-
ticipants'early lives were better remembered than events
within the five years prior to this assessment. No significant
differences were found between recall of childhood and
adulthood events (mean difference¼.16,P¼1.00).
3.1. Findings at ‘baseline’
Fig. 1shows how individual confabulating and non-
confabulating brain-injured patients as well as healthy con-
trols performed on Schnider's TCC task at baseline in the
current study. All the healthy controls scored belowSchnider
et al.'s (1996)cut-off of .28 on this task. 17 of the confabulating
patients scored above this cut-off point, but seven scored
below the cut-off. The non-confabulating brain injured pa-
tients overlapped with both the confabulators and the healthy
controls, five scoring above and six below the cut-off.
One-way ANOVA results showed a near-significant differ-
ence in performance across the three groups of participants
(F
2,38¼3.11;P¼.06).
Table 3compares the performance of the confabulating
and non-confabulating brain-injured groups on TCC and on
our insight, elated mood, and depressed mood scales. There
was no significant difference between these two groups, taken
in isolation, on the TCC scale (t
1,33¼1.28,P¼.21, two-tailed).
They did differ on the insight scale, the confabulating group
showing the lesser insight (t
1,32¼2.26,P¼.03, two-tailed).
They also differed significantly in terms of elated mood, the
confabulating group showing the more elated mood
(t
1,32¼2.05,P¼.04, two-tailed). However, they did not differ
significantly in terms of depressed mood (t
1,32¼.02,P¼.99,
two-tailed).
A backward stepwise binary logistic regression, used to
examine the concurrent associates of confabulation, showed
that for each 1 unit reduction in insight scores (improvement
in insight) there were 30% lower odds of confabulating
Table 2eBackground cognitive tests. Mean raw scores and percentiles of 24 confabulating and 11 non-confabulating brain
injured participants on background cognitive tests.
Test Confabulating participants Non-confabulating participants
Mean (SD)
raw score
Percentile/Classification Mean (SD)
raw score
Percentile/Classification
General reasoning
NART premorbid IQ 93.8 (15.85) 34 90.6 (7.75) 27
WASI current full scale IQ 77.7 (12.70) 7 76.1 (10.78) 5
Speed of processing
SCOLP 22.8 (10.64) 1st e5th 27.3 (8.78) 1st e5th
WAIS-IIIeDigit symbol 23.5 (13.82) 1st 23.4 (13.91) 1st
Executive function
Trail making testeB 254.1 (115.6) 10th 255.0 (130.9) 10th
Cognitive estimates test 14.2 (5.5) 1st 11.6 (5.7) 1st
Hayling testePart A timing score 36.5 (22.6) 5th 30.4 (28.5) 5th
Hayling TestePart B timing score 81.4 (47.5) 10th 80.0 (43.0) 10th
Hayling TestePart C type a&b errors 40.7 (23.6) 1st 30.6 (23.3) 1st
Brixton test 31.1 (10.8) 1st 28.6 (11.1) 1st
Anterograde memory
WMS-IIIeImmediate Logical Memory 12.3 (8.2) 1st 15.3 (6.8) 1st
WMS-IIIeDelayed logical memory 3.1 (4.3) 1st 4.6 (4.7) 1st
WMS-IIIeRecognition logical memory 18.6 (3.6) e 20.5 (2.2) e
WMS-III - Immediate visual reproduction 41.0 (16.7) 1st 49.1 (27.5) 1st
WMS-IIIeDelayed visual reproduction 6.8 (8.9) 2nd 11.5 (13.6) 2nd
WMS-IIIeRecognition visual reproduction 34.6 (4.0) 5th 34.7 (7.8) 5th
Autobiographical memory
AMIeepisodicechildhood 5.0 (.5) Borderline 6.0 (.7) Acceptable
AMIeepisodiceyoung adulthood 4.9 (.4) Borderline 6.4 (.7) Borderline
AMIeepisodicerecent events 3.0 (.5) Abnormal 5.4 (1.0) Abnormal
1
Maulchy's test indicated that the assumption of sphericity
had been violated [c
2
(2)¼.735,P¼.008], therefore degrees of
freedom were corrected. Sinceε>.75, Huynh-Feldt estimates of
sphericity were used to correct degrees of freedom on this occa-
sion (ε¼.851).
cortex 87 (2017) 31 e4336

7. (p¼.03); and for each 1 unit reduction in elated mood scores
there were 33% lower odds of confabulating (p¼.04).Fig. 2
shows a ROC curve for a joint variable (insight, elated mood)
calculated from this regression analysis ( fitted
values¼2.77þ[.36insight score]þ[.40elation score]). In
differentiating confabulating from non-confabulating brain-
injured patients, this ROC curve gave 78% sensitivity and 82%
specificity.
Finally,Table 4shows the correlates of the severity of
confabulation (on the Dalla Barba battery) within the confab-
ulating group. Correlations are shown for the total number of
confabulations across the whole battery, and for the number
of confabulations in the ‘episodic memory’section only. It will
be seen that there was a near-significant correlation between
TCC score and total confabulations (r¼.34,P¼.05) but not
between TCC and episodic confabulations. Significant
correlations were obtained between the number of correla-
tions and AMI episodic memory scores for early adult and
recent life, Logical Memory immediate and delayed, and
cognitive estimates and Hayling test error scores. A stepwise,
forward regression, based on AMI ‘recent’score, TCC scores,
and AMI ‘young adult’score accounted for 49% of the variance
(R
2
) in total confabulation scores.
3.2. Follow-up of confabulation
Recovery from confabulation was examined over a 9-month
period, starting from ‘baseline’, which was actually 6
months after the onset of confabulation in this group (see
Table 1above).
Fig. 3A shows the findings for the confabulating group in
terms of the mean number of confabulations expressed as a
Fig. 1eTemporal Context Confusion (TCC). Scatter distributions of TCC scores for 24 confabulating patients, 11 non-
confabulating brain injured amnesic patients, and 6 healthy controls. TCC¼(FP2/hits2)¡(FP1/hits1). The horizontal line
shows the cut-off point reported for healthy controls (.28) bySchnider et al. (1996)
Table 3eExperimental tasks. Means and standard deviations of scores obtained by each group of participants on
experimental measures, and significance of differences between confabulating and non-confabulating participants
(TCC:Temporal Context Confusion).
Test Confabulating
participants
Non-confabulating
participants
Healthy
participants
Prediction of confabulation in
confabulating versus
Non-confabulating participants
Mean (SD) Mean (SD) Mean (SD)
TCC .5 (.4) .3 (.4) .1 (.1) p¼.21
Insight 8.3 (3.4) 5.6 (2.5) e p¼.03
Elated mood 4.0 (3.3) 1.7 (2.3) 3.2 (3.9) p¼.04
Depression 3.4 (2.8) 3.5 (3.2) 3.5 (3.5) P¼.99
cortex 87 (2017) 31 e43 37

8. proportion of total statements on the Dalla Barba battery, and
also as a proportion of statements on the episodic memory
component of the battery. Both measures showed a steep
decline over 9 months in terms of the proportion of confabu-
latory statements.Fig. 3A also shows the TCC curve from
Schnider's test, superimposed upon the histograms, indi-
cating that TCC appear to have declined in parallel.Fig. 3B
shows changes in mood and insight (high scores ¼poor
insight) over the same time-period. It will be seen that insight
and elated mood improved, and depressed mood showed an
inverted U-shaped curve. Repeated measures ANOVAs
showed a significant linear effect of time on episodic confab-
ulations (F
1,23¼88.78,P¼.00), total confabulations across all
sections of the Dalla Barba confabulation interview
(F
1,22¼85.59,P¼.00), and on insight scores (F
1,21¼18.71,
P¼.00). Similar analyses carried out with the remaining
experimental variables were not significant: TCC
(F
1.37,30.16
2 ¼2.10,P¼.14), depression (c
2
2
¼.34,P¼.86)
3
,or
elation (c
2
2
¼2.70,P¼.26).
3
This indicated there was no sig-
nificant change over time in terms of TCC, elated or depressed
mood.
Next, we examined the correlates of decline in confabula-
tion across individuals by examining the change in total
confabulations on the Dalla Barba battery between baseline
(T1) and 9 months (T3) against change scores (T1eT3) on our
independent variables over the same time-period. We also
examined these correlations for the ‘episodic memory’sec-
tion of the Dalla Barba battery, taken in isolation.Table 5A
Fig. 2eReceiver Operating Characteristics curve of fitted values of insight and elated mood at initial assessment (the dashed
line represents the fitted values of the joint variable (insight&elated mood) and the single diagonal line represents chance
performance).
Table 4eCorrelates of severity in confabulation (within
the confabulating group). Size and significance of
correlations between neuropsychological tests and
number of confabulatory answers to the Dalla Barba
confabulation interview (1. overall and 2. episodic section)
at the first assessment. (TCC: Temporal Context
Confusion; AMI: Autobiographical Memory Interview).
Total
confabulations
T1
Episodic
confabulations
T1
Schnider's TCC score .34 (p¼.05) .01 (P¼.48)
AMI episodic adult .45(p¼.02) .42(p¼.02)
AMI episodic recent .48(p¼.01) .49(p¼.01)
Immediate logical memory .41(p¼.03) .35 (P¼.06)
Delayed logical memory .29 (P¼.10) .40(p¼.03)
Cognitive estimates ERRORS .43 (p¼.02) .37(p¼.04)
Hayling errors .49 (p¼.01) .50(p¼.01)
2
Maulchy's test indicated that the assumption of sphericity
had been violated [c
2
(2)¼12.89,P¼.002], therefore degrees of
freedom were corrected. Greenhouse-Geisser estimates of sphe-
ricity were used to correct degrees of freedom on this occasion
(ε¼.686).
3
The Friedman test for multiple comparisons was used.
cortex 87 (2017) 31 e4338

9. shows that none of the correlations between change in con-
fabulations (T1eT3) and change in TCC score, elated mood,
depressed mood, and insight score was statistically significant
(r¼.05e.26). As shown inTable 5B, it was change in memory
scores and change in cognitive estimates error scores (T1eT3)
which were statistically significant. A stepwise regression,
based on change in AMI episodic young adult score and
change in cognitive estimates error scores predicted 44% of
the variance (R
2
) of change in total confabulation scores over 9
months.
0
.1
.2
.3
.4
.5
.6
0
.1
.2
.3
.4
.5
.6
.7
ini?al assessment 3-month follow-
up
9-month follow-
up
Temporal Context Confusion
episodic confabula?ons
overall confab
episodic confab
TCC
.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
ini?al
assessment
3-month
follow-up
9-month
follow-up
mood & insight scores
% confabula?ons
episodic confab
insight
elated mood
depressed mood
A
B
Fig. 3eA&B. Confabulation and experimental measures over time. A. Clinical course of confabulation and Temporal
Context Confusion (TCC) over time. 1. Left vertical axis: Mean proportion of confabulatory answers reported on the Dalla
Barba interview a) across all sections, or b) on the episodic section alone; and 2. Right vertical axis: Mean TCC scores of 24
confabulating participants at 3 assessment times. Lower scores indicate better TCC performance. B. Episodic
confabulations, insight, elation and depressed mood over time. Left vertical axis: raw number of confabulatory responses to
the Dalla Barba episodic section over time. Right vertical axis: Raw scores obtained by the confabulating group over time on
a) insight, b) elation, and c) depression questionnaires. Decreasing scores indicate better functioning.
Table 5AeCorrelates ofdeclinein confabulation across 24
confabulating individuals on A. Experimental measures.
Correlation values between change in experimental
measures and change in episodic/overall confabulatory
answers (change in this case is measured by subtracting
final assessment scores from initial scores).
Total
confabulations
T1eT3
Episodic
confabulations
T1eT3
Temporal context
confusion T1eT3
.05 (p¼.41) .02 ( p¼.46)
Elated mood T1eT3 .10 (p¼.32) .07 (p¼.38)
Depressed mood T1eT3 .26 (p¼.12) .25 ( p¼.13)
Insight .10 ( p¼.33) .09 ( p¼.34)
Table 5BeCorrelates ofdeclinein confabulation across 24
confabulating individuals on B. Background cognitive
measures. Size and significance of correlations between
change in neuropsychology measures and change in
episodic/overall confabulatory answers (change in this
case is measured by subtracting scores final assessment
from initial scores) (AMI: Autobiographical Memory
Interview; T1: baseline; T3: final assessment).
Total
confabulations
T1eT3
Episodic
confabulations
T1eT3
AMI episodic child T1eT3 .43 (p¼.035) .06 (p¼.40)
AMI episodic adult T1eT3 .63(p¼.00) .31 (p¼.09)
Delayed logical
memory T1eT3
.26 (p¼.13) .40(p¼.03)
Cognitive estimates
errors T1eT3
.46(p¼.02) .51(p¼.01)
cortex 87 (2017) 31 e43 39

10. 4. Discussion
4.1. Confabulating versus non-confabulating groups
In this study, we compared the performance of confabulating
versus non-confabulating brain-injured patients with healthy
controls. Confabulation was determined on clinical grounds
and by scores on the episodic component of Dalla Barba 's
battery. The clinical and Dalla Barba criteria were entirely
consistent, and none of the patients had to be excluded
because their Dalla Barba scores did not match their clinical
diagnosis. This makes it very unlikely that the phenomenon
was purely a consequence of the nature of the interview or
discussion between the experimenter and the participant
(compareMcVittie et al., 2014). Furthermore, 21 out of 24 of the
confabulators had acted upon their confabulations, consistent
withSchnider's (2003, 2008)more specific definition of ‘spon-
taneous confabulation’.
Both groups of patients performed similarly on background
neuropsychological tests with a decline in estimated current
I.Q., relative to estimated premorbid scores, and impairments
on tests of anterograde memory, executive function, and
speed of processing.
On the AMI, both groups performed significantly worse in
terms of ‘recent’personal memories, compared with child-
hood or young adult memories. There was also a main effect
of group, indicating that the confabulating patients recalled
autobiographical events less accurately than did the non-
confabulating memory-disordered patients. Although there
was not any significant group by timeeperiod interaction, the
findings are consistent with the suggestion that older auto-
biographical memories are more stable (Dalla Barba, 2009;
Dalla Barba&Boisse, 2010; Dalla Barba et al., 1997).
4.2. Findings at baseline
4.2.1. The role of TCC in the onset of confabulation
Schnider and colleagues have claimed that TCC ( Schnider
et al., 2000) or reality monitoring (Nahum et al., 2012) is the
only correlate of both the presence and clinical course of
confabulation (Manuel et al., 2014; Schnider et al., 2013, 2000).
TCC is defined as the inability to filter out memories that do
not belong to ongoing reality (Schnider, 2003). Results from the
current study partially replicated the work by Schnider and
colleagues. Seventy one percent of the present confabulators
scored above the cut-off point that would index confabulation
according to Schnider and colleagues, and all healthy controls
scored below this point. In this way the present results
constitute a partial replication of Schnider's previous findings
(Schnider et al., 1996).
However, contrary to Schnider's findings, seven confabu-
lating participants scored below the critical cut-off point. In
addition, the scores of the non-confabulating brain-injured
patients were evenly distributed across both sides of the
critical cut-off point (Fig. 1). This indicated that performance
on the TCC task did not always discriminate between
confabulating and non-confabulating memory-disordered
patients. The non-confabulating brain injured participants'
performance on this task was very similar to that of the non-
confabulating ACoA group reported byGilboa et al. (2006). This
may relate to the severity of executive dysfunction experi-
enced by both our brain-injured and Gilboa et al.’s ACoA
groups. Moreover, TCC did not differentiate significantly the
confabulating and non-confabulating brain-injured groups;
and it correlated only at a near-significant level (p<.06) with
total confabulations on the Dalla Barba battery.
4.2.2. The role of insight and mood state in the onset of
confabulation
Our measures of ‘insight’and ‘elated mood’did each differ
significantly between the confabulating and non-
confabulating brain-injured groups. Moreover, a ROC anal-
ysis, based on these two variables, predicted group member-
ship (confabulating versus non-confabulating brain injury)
with 78% sensitivity and 82% specificity.
The role of insight in the onset and clinical course of
confabulation has not been formally studied previously. The
present findings provided evidence that insight into memory
difficulties and well-being might be an important factor in
the appearance and clinical course of confabulation. These
results are consistent with previous observations indicating
that recovery from confabulation is accompanied by
improved awareness of deficits (Feinberg&Roane, 1997;
Mercer et al., 1977). Somewhat similarly,Schnider (2008)
has argued that confabulation is caused by deficits in expe-
riencing ongoing reality, and others have cited a disconnec-
tion from the external world (Feinberg&Roane, 1997), or a
lack of integration in memory processes (Conway, Collins,
Gathercole,&Anderson, 1996). Indeed, there is some evi-
dence that clinical interventions, aimed at improving
awareness and self-monitoring have improved confabulation
in clinical patients, although the precise mechanism whereby
this has happened is not clear (Dayus&Van den Broek, 2000;
DeLuca, 2009). With respect to mood state, the present find-
ings are consistent with previous observations that elated
mood is associated with a liberal response bias on recogni-
tion memory tests, which has been reported to subside as the
patient's mood returns to normal (Corwin, et al., 1990). Such
findings are consistent with other evidence that affective
factors may influence the onset and characteristics of
confabulation in patients with weak memory or impaired
monitoring mechanisms (Conway, 2005; Fotopoulou, 2010;
Gilboa, 2010; Metcalf et al., 2010). In a previous study, our
group found that confabulated memories contained a
significantly higher proportion of affective content (both
pleasant and unpleasant) than did ‘true’memories (Bajo
et al., 2010). It is conceivable that these affective factors
might influence strategic control, trace specification, or
verification processes in memory retrieval ( Burgess&
Shallice, 1996; Gilboa et al., 2006; Moscovitch&Melo, 1997;
Schacter et al., 1998).
4.2.3. Memory and executive function in the onset of
confabulation
The main factors which correlated with the severity of the
total and episodic confabulations across individual patients
were impaired autobiographical memory (young adult and
recent memories), verbal anterograde memory (immediate
and delayed Logical Memory), and errors on executive tests
cortex 87 (2017) 31 e4340

11. (cognitive estimates, Hayling test). Taken together, a stepwise
regression, based on AMI young adult and recent scores, as
well as the TCC score, predicted 49% of the variance in terms
of total confabulations. Previous studies have obtained vari-
able findings with respect to the association between execu-
tive test performance and confabulation, some reporting
strong associations (e.g.,DeLuca&Cicerone, 1991; DeLuca&
Locker, 1996; Ghosh et al., 2014; Kapur&Coughlan, 1980;
Stuss et al., 1978) and others reporting weak or no associa-
tion (Bouzerda-Wahlen et al., 2013; Gilboa, et al., 2006).
Johnson et al. (1997)suggested that a weak autobiographical
memory was also a factor promoting confabulation on the
basis of a single case but, to our knowledge, this has not been
systematically demonstrated before.
4.3. Follow-up of confabulation
Very few previous studies have examined the clinical course
of confabulation, and its correlates.Schnider et al. (2000),
Nahum et al. (2012, 2010)found that TCC/reality disorienta-
tion did indeed decline in parallel with recovery from
confabulation.Box, Laing, and Kopelman (1999)showed that
confabulation can occasionally evolve into other types of
delusional disorder.Schnider (2008)reported that dopamine
antagonists sometimes reduce confabulation.Corwin et al.
(1990)found that beta-blockers reverse elated mood disor-
ders, and can dampen an excessive willingness to guess dur-
ing recognition memory tasks.
In the present investigation, we found that, across the
entire group, TCC errors did indeed come down in parallel
with the decline in confabulations, asSchnider et al. (2000)
had previously demonstrated. On the other hand, there were
no significant correlations between the decline in TCC (be-
tween T1 and T3) with either the total number or purely
episodic confabulations on the Dalla Barba battery.
Likewise, our measure of the loss of insight also improved
through time in parallel with recovery from confabulation;
and there was some improvement in terms of elated mood,
again in parallel with recovery from confabulation. However,
neither of these measures (nor depressed mood) was signifi-
cantly associated with either the decline in total confabula-
tion or the decline in episodic confabulation scores.
What appeared to be significant correlates of the decline in
confabulation were improvements in AMI episodic child and
young adult memories, verbal anterograde memory (delayed
Logical Memory scores) and, in particular, the decline in errors
on the CET. This suggests the importance of memory perfor-
mance (particularly autobiographical memory) and improve-
ments in executive function in determining the outcome of
confabulation. Stepwise regression based on change in
cognitive estimates and change in AMI scores predicted 44% of
the variance in terms of change in total confabulation scores.
Again, this finding is consistent with the observations of some
of the older literature, which placed emphasis on a combi-
nation of executive and memory dysfunction ( Baddeley&
Wilson, 1986; DeLuca&Cicerone, 1991; Kopelman, 1987). It
highlights the inability to edit out errors, as demonstrated by
Shallice and Evans (1978), and a weak autobiographical
memory, as postulated byJohnson et al. (1997).
4.4. Strengths and limitations
The key strength of this study was the relatively large number
of participants, compared with previous investigations, who
are representative of brain injured patients seen in routine
NHS practice. Another strength was that they were followed
up for 9 months. The main limitation was the variation in
stage of development of the different schedules we have used.
The newest schedules such as the insight measure were likely
to be subject to a greater degree of measurement error, which
may have partly obscured any association with confabulation,
and consequently comparisons between the schedules in the
strength of their association with confabulation must be made
with caution.
4.5. Summary and conclusions
In summary, TCC reality orientation was sensitive, but not
specific, to confabulation. There was a weak correlation at
baseline with the severity of total confabulations across pa-
tients, although this was not present for episodic confabula-
tions, taken in isolation. Elation and low insight better
predicted the presence of confabulation versus no confabu-
lation in brain injury patients, although these did not correlate
significantly with the severity of confabulation, and they
could conceivably be a consequence (rather than the cause) of
confabulation. Significant correlates of episodic and total
confabulations were executive function (errors on the Hayling
and CETs), poor anterograde memory (story recall), and weak
autobiographical memory retrieval (on the AMI).
With respect to change through time, TCC, insight score
and elation score, all improved in parallel with confabulation.
However, these factors were not significantly correlated with
change in confabulation scores at 9-months’follow-up. From
our correlational analyses, it would appear that weak memory
(especially autobiographical memory) and executive function
(as measured by cognitive estimate scores) were the strongest
correlates ofboththe initial severity of confabulationand
change in confabulation over 9 months.
More generally, elated mood and lack of insight may
contribute to the occurrence of confabulation by creating
excessive confidence and willingness to report false mem-
ories. TCC/impaired reality monitoring (at least as measured
by the Schnider test) is often present in confabulation, and in
general decline in parallel with recovery from confabulation,
but these confusions are not specific to the presence of
confabulation nor closely correlated with its severity. It ap-
pears that spontaneous confabulation is most likely to arise in
the context of impaired autobiographical memory and exec-
utive function, particularly with respect to the monitoring/
editing out of errors. This failure to edit out errors may also
include the failure to reject inappropriate schema, as sug-
gested byGhosh et al. (2014).
Acknowledgements
We would like to thank all participants and their carers who
helped with this research. We are also grateful for the support
cortex 87 (2017) 31 e43 41

12. offered by the following NHS services: Brain Injury Rehabili-
tation Unit (BIRU, Edgware Hospital), Lishman Brain Injury
Unit (Maudsley hospital), St Thomas'Hospital Memory Clinic
(Psychological Medicine) and Blackheath Brain Injury Unit.
Supplementary data
Supplementary data related to this article can be found at
http://dx.doi.org/10.1016/j.cortex.2016.06.016.
Funding
None.
references
Altman, E. G., Hedeker, D., Peterson, J. L., & Davis, J. M. (1997). The
altman self-rating Mania scale.Biological Psychiatry, 42(10),
948e955.
Baddeley, A. D., Emslie, H., & Nimmo-Smith, I. (1992).The speed
and capacity of language-processing testeManual. Thames Valley
Test Company Ltd.
Baddeley, A., & Wilson, B. (1986). Amnesia, autobiographical
memory, and confabulation. In C. David, & Rubin (Eds.),
Autobiographical memory(pp. 225e252). New York, NY:
Cambridge University Press; US.
Bajo, A., Fleminger, S., & Kopelman, M. (2010). Confabulations are
emotionally charged, but not always for the best.Journal of the
International Neuropsychological Society, 16(06), 975e983.http://
dx.doi.org/10.1017/S1355617710000913.
Bouzerda-Wahlen, A., Nahum, L., Ptak, R., & Schnider, A. (2013).
Mechanism of disorientation: Reality filtering versus content
monitoring.Cortex, 49(10), 2628e2636.
Box, O., Laing, H., & Kopelman, M. (1999). The evolution of
spontaneous confabulation, delusional misidentification and
a related delusion in a case of severe head injury.Neurocase,
5(3), 251e262.
Burgess, P. W., & Shallice, T. (1996). Confabulation and the control
of recollection.Memory, 4(4), 359e411.
Burgess, P. W., & Shallice, T. (1997).The Hayling and Brixton Testse
Manual Bury St Edmunds. UK: Thames Valley Test Company
Ltd.
Conway, M. A. (2005). Memory and the self.Journal of Memory and
Language, 53(4), 594e628.
Conway, M. A., Collins, A. F., Gathercole, S. E., & Anderson, S. J.
(1996). Recollections of true and false autobiographical
memories.Journal of Experimental Psychology: General, 125(1),
69e95.
Conway, M. A., & Tacchi, P. C. (1996). Motivated confabulation.
Neurocase, 2(4), 325e338.
Corwin, J., Peselow, E., Feenan, K., Rotrosen, J., & Fieve, R. (1990).
Disorders of decision in affective disease: An effect of beta-
adrenergic dysfunction?Biological Psychiatry, 27(8), 813e833.
Dalla Barba, G. (1993). Confabulation: Knowledge and recollective
experience.Cognitive Neuropsychology, 10(1), 1e20.
Dalla Barba, G. (2009). Temporal consciousness and
confabulation: Escape from unconscious explanatory idols. In
W. Hirstein (Ed.),Confabulation. views from neuroscience,
psychiatry, psychology, and philosophy(pp. 223e262). New York:
Oxford University Press (Reprinted from: NOT IN FILE).
Dalla Barba, G., & Boisse, M. F. (2010). Temporal consciousness
and confabulation: Is the medial temporal lobe“temporal”?
Cognitive Neuropsychiatry, 15(1), 95e117.
Dalla Barba, G., Cappelletti, J. Y., Signorini, M., & Denes, G. (1997).
Confabulation: Remembering ‘another’past, planning
‘another’future.Neurocase, 3(6), 425e435.
Dayus, B., & Van den Broek, M. D. (2000). Treatment of stable
delusional confabulations using self-monitoring training.
Neuropsychological Rehabilitation, 10(4), 415e427.
DeLuca, J. (2000). A cognitive neuroscience perspective on
confabulation.Neuro-Psychoanalysis, 2(2), 119e132.
DeLuca, J. (2009). Confabulation in anterior communicating artery
syndrome. In W. Hirstein (Ed.),Confabulation. Views from
neuroscience, psychiatry, psychology, and philosophy(pp. 13e32).
New York: Oxford University Press (Reprinted from: NOT IN
FILE).
DeLuca, J., & Cicerone, K. D. (1991). Confabulation following
aneurysm of the anterior communicating artery.Cortex, 27(3),
417e423.
DeLuca, J., & Locker, R. (1996). Cognitive rehabilitation following
anterior communicating artery aneurysm bleeding: A case
report [Case Reports Research Support, Non-U.S. Gov't]
Disability and Rehabilitation, 18(5), 265e272.
Feinberg, T. E., & Roane, D. M. (1997). Anosognosia, completion
and confabulation: The neutral-personal dichotomy.
Neurocase, 3(1), 73e85.
Fotopoulou, A. (2008). False selves in neuropsychological
rehabilitation: The challenge of confabulation.
Neuropsychological Rehabilitation, 18(5e6), 541e565.
Fotopoulou, A. (2010). The affective neuropsychology of
confabulation and delusion.Cognitive Neuropsychiatry, 15(1),
38e63.
Fotopoulou, A., Conway, M., Griffiths, P., Birchall, D., & Tyrer, S.
(2007). Self-enhancing confabulation: Revisiting the
motivational hypothesis.Neurocase, 13(1), 6e15
.
Fotopoulou, A., Conway, M. A., Solms, M., Tyrer, S., &
Kopelman, M. (2008a). Self-serving confabulation in prose
recall.Neuropsychologia, 46(5), 1429e1441.
Fotopoulou, A., Conway, M. A., Tyrer, S., Birchall, D., Griffiths, P.,
& Solms, M. (2008b). Is the content of confabulation positive?
An experimental study.Cortex, 44(7), 764e772.
Fotopoulou, A., Solms, M., & Turnbull, O. (2004). Wishful reality
distortions in confabulation: A case report.Neuropsychologia,
42(6), 727e744.
Gainotti, G., & Marra, C. (1994). Some aspects of memory
disorders clearly distinguish dementia of the Alzheimer's type
from depressive pseudo-dementia.Journal of Clinical&
Experimental Neuropsychology, 16(1), 65e78.
Ghosh, V. E., & Gilboa, A. (2014). What is a memory schema? A
historical perspective on current neuroscience literature.
[Review].Neuropsychologia, 53(1), 104e114.
Ghosh, V. E., Moscovitch, M., Colella, M.,B., & Gilboa, A. (2014).
Schema representation in patients with ventromedial PFC
lesions.The Journal of Neuroscience, 34(36), 12057e12070.
Gilboa, A. (2010). Strategic retrieval, confabulations, and
delusions: Theory and data.Cognitive Neuropsychiatry, 15(1),
145e180.
Gilboa, A., Alain, C., Stuss, D. T., Melo, B., Miller, S., &
Moscovitch, M. (2006). Mechanisms of spontaneous
confabulations: A strategic retrieval account.Brain, 129(6),
1399e1414.
Gilboa, A., & Moscovitch, M. (2002). The cognitive neuroscience of
confabulation: A review and a model. In A. Baddeley,
M. D. Kopelman, & B. A. Wilson (Eds.),The handbook of memory
disorders(Vol. 2nd ed., pp. 315e342). Chichester, UK: John
Wiley&Sons Ltd.
Gilboa, A., & Verfaellie, M. (2010). Introduction-telling it like it
isn't: The cognitive neuroscience of confabulation.Journal of
cortex 87 (2017) 31 e4342

13. the International Neuropsychological Society, 16(06), 961e966.
http://dx.doi.org/10.1017/S135561771000113X.
Hebscher, M., Barkan-Abramski, M., Goldsmith, M., Aharon-
Peretz, J., & Gilboa, A. (2015). Memory, decision-making, and
the ventromedial prefrontal cortex (vmPFC): The roles of
subcallosal and posterior orbitofrontal cortices in monitoring
and control processes.Cerebral Cortex.http://dx.doi.org/
10.1093/cercor/bhv220. bhv220 [pii].
Johnson, M. K., O'Connor, M., & Cantor, J. (1997). Confabulation,
memory deficits, and frontal dysfunction.Brain and Cognition,
34(2), 189e206.
Kapur, N., & Coughlan, A. K. (1980). Confabulation and frontal
lobe dysfunction.Journal of Neurology, Neurosurgery&
Psychiatry, 43(5), 461e463.
Kopelman, M. D. (1987). Two types of confabulation.Journal of
Neurology Neurosurgery and Psychiatry, 50(11), 1482e1487.
Kopelman, M. D. (1999). Varieties of false memory.Cognitive
Neuropsychology, 16(3e5), 197e214.
Kopelman, M. D. (2010). Varieties of confabulation and delusion.
Cognitive Neuropsychiatry, 15(1), 14e37.
Kopelman, M. D., Ng, N., & Van Den Brouke, O. (1997).
Confabulation extending across episodic, personal, and
general semantic memory.Cognitive Neuropsychology, 14(5),
683e712.
Kopelman, M. D., Wilson, B. A., & Baddeley, A. (1990).The
autobiographical memory interview Bury St Edmunds. UK: Thames
Valley Test Company Ltd.
La Corte, V., Serra, M., Attali, E., Boisse, M. F., & Dalla Barba, G.
(2010). Confabulation in Alzheimer's disease and amnesia: A
qualitative account and a new taxonomy.Journal of the
International Neuropsychological Society, 16(06), 967e974.http://
dx.doi.org/10.1017/S1355617710001001.
Luria, A. R. (1976).The neuropsychology of memory. (Trans B. Haigh).
Oxford, England: V. H. Winston&Sons; England.
Manuel, A. L., David, A. W., Bikson, M., & Schnider, A. (2014).
Frontal tDCS modulates orbitofrontal reality filtering
[Research Support, Non-U.S. Gov't]Neuroscience, 265,21e27.
McVittie, C., McKinlay, A., Della Sala, S., & MacPherson, S. E.
(2014). The dog that didn't growl: The interactional negotiation
of momentary confabulations.Memory, 22(7), 824e838.
Mercer, B., Wapner, W., Gardner, H., & Benson, D. F. (1977). A
study of confabulation.Archives of Neurology, 34(7), 429e433.
Metcalf, K., Langdon, R., & Coltheart, M. (2010). The role of
personal biases in the explanation of confabulation.Cognitive
Neuropsychiatry, 15(1), 64e94.
Moscovitch, M., & Melo, B. (1997). Strategic retrieval and the
frontal lobes: Evidence from confabulation and amnesia.
Neuropsychologia, 35(7), 1017e1034.
Nahum, L., Bouzerda-Wahlen, A., Guggisberg, A., Ptak, R., &
Schnider, A. (2012). Forms of confabulation: Dissociations and
associations.Neuropsychologia, 50(10), 2524e2534.
Nahum, L., Ptak, R., Leemann, B., Lalive, P., & Schnider, A. (2010).
Behaviorally spontaneous confabulation in limbic
encephalitis: The roles of reality filtering and strategic
monitoring.Journal of the International Neuropsychological
Society, 16(06), 995e1005.http://dx.doi.org/10.1017/
S1355617710000780.
Nelson, H., & Willison, J. (1991).National adult reading test (NART)
second edition. Test manual (Vol. Second Edition). Windsor: NFER-
NELSON Publishing Company Ltd, UK .
Reitan, R. M. (1958). Validity of the trail making test as an
indicator of organic brain damage.Perceptual and Motor Skills, 8,
271e276.
Schacter, D. L., Norman, K. A., & Koutstaal, W. (1998). The
cognitive neuroscience of constructive memory [Review] [157
refs]Annual Review of Psychology, 49, 289e318.
Schnider, A. (2001). Spontaneous confabulation, reality
monitoring, and the limbic systemea review [Review] [70 refs]
Brain ResearcheBrain Research Reviews, 36(2e3), 150e160.
Schnider, A. (2003). Spontaneous confabulation and the
adaptation of thought to ongoing reality [Review] [101 refs]
Nature Reviews Neuroscience, 4(8), 662e671.
Schnider, A. (2008).The confabulating mind. How the brain creates
reality. New York: Oxford university press.
Schnider, A., Nahum, L., Pignat, J. M., Leemann, B., Lovblad, K. O.,
Wissmeyer, M., et al. (2013). Isolated prospective
confabulation in Wernicke-Korsakoff syndrome: A case for
reality filtering.Neurocase, 19
(1), 90e104.
Schnider, A., & Ptak, R. (1999). Spontaneous confabulators fail to
suppress currently irrelevant memory traces [see comment]
Nature Neuroscience, 2(7), 677e681.
Schnider, A., Ptak, R., von Daniken, C., & Remonda, L. (2000).
Recovery from spontaneous confabulations parallels recovery
of temporal confusion in memory.Neurology, 55(1), 74e83.
Schnider, A., von Daniken, C., & Gutbrod, K. (1996). The
mechanisms of spontaneous and provoked confabulations.
Brain, 119(Pt 4), 1365e1375.
Shallice, T., & Evans, M. E. (1978). The involvement of the frontal
lobes in cognitive estimation.Cortex, 14(2), 294e303.
Snodgrass, J. G., & Vanderwart, M. (1980). A standardized set of
260 pictures: Norms for name agreement, image agreement,
familiarity, and visual complexity.Journal of Experimental
Psychology: Human Learning&Memory, 6(2), 174e215.
Stuss, D. T., Alexander, M. P., Lieberman, A., & Levine, H. (1978).
An extraordinary form of confabulation.Neurology, 28(11),
1166e1172.
Toosy, A. T., Burbridge, S. E., Pitkanen, M., Loyal, A. S., Akanuma, N.,
Laing, H., et al. (2008). Functional imaging correlates of fronto-
temporal dysfunction in Morvan's syndrome.Journal of
Neurology, Neurosurgery&Psychiatry, 79(6), 734e735.
Turnbull, O. H., Berry, H., & Evans, C. E. (2004). A positive
emotional bias in confabulatory false beliefs about place
[Review] [47 refs]Brain and Cognition, 55(3), 490e494.
Turnbull, O., Jenkins, S., & Rowley, M. L. (2004). The pleasantness
of false beliefs: An emotion-based account of confabulation.
Neuro-Psychoanalysis, 6(1), 5e16.
Turner, M. S., Cipolotti, L., Yousry, T. A., & Shallice, T. (2008).
Confabulation: Damage to a specific inferior medial prefrontal
system.Cortex, 44(6), 637e648.
Walsh, B. (2004).Multiple comparisons: Bonferroni corrections and
false discovery rates, 32676. fromhttp://nitro.biosci.arizona.
edu/workshops/Aarhus2006/pdfs/Multiple.pdf.
Wechsler, D. (1999).WASI. Wechsler Abbreviated Scale of Intelligence.
Manual London: The Psychological Corporation Ltd.
Wechsler, D., Wycherley, R. J., Benjamin, L., Crawford, J., &
Mockler, D. (1997a). WAIS-III UK. Wechsler adult intelligence
scale. InAdministration and scoring manual(3rd ed., Vol. 3).
London: The Psychological Corporation Ltd
.
Wechsler, D., Wycherley, R. J., Benjamin, L., Crawford, J., &
Mockler, D. (1997b). WMS-III UK. Wechsler memory scale. In
Administration and scoring manual(3rd ed., Vol. 3). London: The
Psychological Corporation Ltd.
Weinstein, E. A., & Kahn, R. L. (1950). The syndrome of anosognosia.
Archives of Neurology and Psychiatry, 64(6), 772e791.
Zigmond, A. S., & Snaith, R. P. (1983). The hospital anxiety and
depression scale.Acta Psychiatrica Scandinavica, 67(6), 361e370.
cortex 87 (2017) 31 e43 43